Obsessive-compulsive disorder (OCD) is an important public health problem because of its 1) high lifetime prevalence rate (2% to 3% in the general population), 2) often chronic course and 3) frequently accompanying interpersonal, social and occupational dysfunction. Neurologic diseases with obsessive-compulsive symptoms, neurosurgical treatments for OCD and previous brain imaging studies have suggested, but not conclusively demonstrated, pathology in the frontal lobes (especially the orbital complex), limbic structures and basal ganglia in OCD. The primary goal of this study is to test the association between these structures and OCD by comparing the volumes of these and associated brain structures in OCD subjects and normal controls. To accomplish this, Magnetic Resonance scans will be obtained from 50 OCD subjects and 50 age, sex, race and socioeconomic status-matched normal control subjects. Using a computerized volumetric mensuration system, the volume of the following brain structures will be assessed, under double blind conditions, on each scan: 1) the premotor, prefrontal and orbital cortical subdivisions as parts of the frontal lobe; 2) the prefrontal and premotor interhemispheric fissure subdivisions (an indirect measure of change in frontal lobe size); 3) limbic structures (cingulate gyrus and amygdala-hippocampus complex); 4) the lateral ventricle with its subdivisions, the frontal horns, central parts, occipital horns and temporal horns (an indirect measure of change in volume of the hemispheres, thalamus and basal ganglia); 5) the basal ganglia (the caudate nucleus, nucleus accumbens and putamen); 6) and whole brain (to determine if changes in volume of the above structures are secondary to generalized brain volume changes or are structure-specific). The assessed volumes of these regions for the two groups will then be compared. If differences are found between the OCD and normal control groups, the specificity of these changes will be assessed by comparing the structure volumes in the OCD group with those of subjects with medication- naive psychosis (N=90), bipolar disorder without psychosis (N=20) and major depression without psychosis (N=20) available from other ongoing funded studies which use the same MR procedures as this proposal.